Real Time Face Mask Detection with TensorFlow and Python

Real Time Face Mask Detection with TensorFlow and Python

Prof. P.G.Patil1  Madhav R.kumar2 Kailas S.Shirole3  Dheeraj G.londhe4

1. Prof. P.G.Patil Electronics and Telecommunication Engineering SAE

2. Madhav R.kumar Electronics and Telecommunication Engineering  SAE

3. Kailas S.Shirole  Electronics and Telecommunication Engineering SAE

4. Dheeraj G.londhe  Electronics and Telecommunication Engineering SAE

---------------------------------------------------------------------***---------------------------------------------------------------------

ABSTRACT :

The COVID-19 pandemic has driven the development of real-time face mask detection systems. This project details a system built with TensorFlow and Python for this purpose. It involves three steps: data collection, model training, and real-time detection. First, a dataset of labeled images (masked/unmasked faces) is prepared. Then, a Convolutional Neural Network (CNN) is trained using TensorFlow and Keras to classify faces. Transfer learning can be used for improved performance. Finally, the trained model is integrated with OpenCV for real-time video processing. Faces are identified in each frame, and the model predicts if a mask is worn. This information can be used for alerts or notifications, promoting mask compliance in public spaces like airports, schools, and public transport. This versatile system offers applications in public health and security, leveraging TensorFlow's deep learning and Python's real-time processing capabilities for disease control and public safety.

Key Words: Real-time face mask detection, Convolutional Neural Network (CNN), TensorFlow & Keras, Transfer learning